Codon optimization and mRNA amplification effectively ... - Nature

Gene Therapy (2004) 11, 522–533 & 2004 Nature Publishing Group All rights reserved 0969-7128/04 $25.00 www.nature.com/gt

RESEARCH ARTICLE

Codon optimization and mRNA amplification effectively enhances the immunogenicity of the hepatitis C virus nonstructural 3/4A gene L Frelin1, G Ahle´n1, M Alheim1, O Weiland2, C Barnfield3, P Liljestro¨m3,4 and M Sa¨llberg1 1

Division of Clinical Virology, , Karolinska Institutet at Huddinge University Hospital, Stockholm, Sweden; 2Division of Infectious Diseases, Karolinska Institutet at Huddinge University Hospital, Stockholm, Sweden; 3Swedish Institute for Infectious Disease Control, Stockholm, Sweden; and 4Microbiology and Tumourbiology Centre, Karolinska Institutet, Stockholm, Sweden

We have recently shown that the NS3-based genetic immunogens should contain also hepatitis C virus (HCV) nonstructural (NS) 4A to utilize fully the immunogenicity of NS3. The next step was to try to enhance immunogenicity by modifying translation or mRNA synthesis. To enhance translation efficiency, a synthetic NS3/4A-based DNA (coNS3/4A-DNA) vaccine was generated in which the codon usage was optimized (co) for human cells. In a second approach, expression of the wild-type (wt) NS3/4A gene was enhanced by mRNA amplification using the Semliki forest virus (SFV) replicon (wtNS3/4A-SFV). Transient tranfections of human HepG2 cells showed that the coNS3/4A gene gave 11-fold higher levels of NS3 as compared to the wtNS3/4A gene when using the CMV promoter. We have previously shown that the presence of NS4A enhances the expression by SFV. Both codon optimization and mRNA amplification resulted in an improved immunogenicity as evidenced by higher levels of NS3-specific antibodies. This improved

immunogenicity also resulted in a more rapid priming of cytotoxic T lymphocytes (CTLs). Since HCV is a noncytolytic virus, the functionality of the primed CTL responses was evaluated by an in vivo challenge with NS3/4A-expressing syngeneic tumor cells. The priming of a tumor protective immunity required an endogenous production of the immunogen and CD8 þ CTLs, but was independent of B and CD4 þ T cells. This model confirmed the more rapid in vivo activation of an NS3/4A-specific tumor-inhibiting immunity by codon optimization and mRNA amplification. Finally, therapeutic vaccination with the coNS3/4A gene using gene gun 6–12 days after injection of tumors significantly reduced the tumor growth in vivo. Codon optimization and mRNA amplification effectively enhances the overall immunogenicity of NS3/4A. Thus, either, or both, of these approaches should be utilized in an NS3/4A-based HCV genetic vaccine. Gene Therapy (2004) 11, 522–533. doi:10.1038/sj.gt.3302184

Keywords: hepatitis C virus; HCV; NS3; DNA vaccine; SFV; codon optimization

Introduction A major problem with chronic hepatitis C virus (HCV) infection is the patients infected by HCV of genotype 1 since these patients respond poorly to existing therapies.1 New therapies for patients infected with HCV of genotype 1 are therefore needed. Most patients infected with HCV develop chronic infections possibly due to the high genetic variability of HCV.2 The genetic heterogeneity of HCV is a result of the high viral replication rate of 1010–1013 viral particles daily3 together with spontaneous nucleotide substitutions of 102–103 substitutions per nucleotide per year.4,5 This suggests that vaccine development should be targeted against genetically stable regions of the HCV genome. It has been shown that DNA immunizations can induce both specific antibodies and cell-mediated responses against the structural and nonstructural (NS) HCV proteins in mice.6–16 Correspondence: M Sa¨llberg, Division of Clinical Virology, F 68, Huddinge University Hospital, S-141 86 Stockholm, Sweden Received 4 July 2003; accepted 26 September 2003

The NS protein 3 of HCV has been considered as a possible vaccine target. The reasons for this are that the NS3 protein shows a limited genetic variability, it performs multiple enzymatic functions, and it is a relatively large protein.17–20 We noted that when using a genetic immunogen containing the complete NS3/4A protease, the humoral responses were surprisingly strong.13 Previous reports had shown that the presence of the cofactor NS4A increases the intracellular stability of NS3 and targets NS3 to intracellular membranes, factors that possibly may affect immunogenicity.21,22 However, we could recently show that the presence of NS4A, by some mechanism, increases the expression levels of NS3, which explains the increased immunogenicity conferred by NS4A.23 Interestingly, a recent report showed that the NS3/4A complex interferes with the interferon signaling pathways in transfected cells.24 Apart from implying an immunomodulatory role in natural infection, this may further affect the NS3 expression in the presence of NS4A observed when using the Semliki forest virus (SFV) vector system.23 In addition to improving expression levels by the inclusion of NS4A or mRNA amplification by the SFV

Codon optimization and RNA amplification for HCV NS3/4A L Frelin et al

523

replicase, the immunogenicity of NS3 may possibly be improved by alternative modifications. One approach, which has been shown to sometimes improve expression levels, is codon optimization. By adjusting the codon usage within the expressed gene to the codons most commonly used by human cells, an increased expression of the gene of interest may be achieved. This is even true for viral genes from viruses that infect humans as the natural host.25,26 One would assume that such viral genes would already be humanized due to the co-evolution with the human host. However, the virus may have for certain reasons evolved not to have all or certain genes codon optimized since this might help the virus to avoid the immune response. Since HCV has one open reading frame from which a large polyprotein is produced, it is unlikely that one single protein gene should have been codon optimized through evolution. Thus, this approach should be evaluated for the ability to enhance expression of HCV proteins. We previously showed that NS4A enhances the immunogenicity of NS3 by increasing expression levels. We have demonstrated here that the immunogenicity of NS3 can be further improved by either codon optimization or the use of mRNA amplification, which has important implications for development NS3-based HCV vaccines.

Results Characterization of the codon-optimized NS3/4A gene and SFV expression vectors The expressions of NS3 and NS3/4A proteins from the wtNS3/4A and coNS3/4A plasmids were analyzed by an in vitro transcription and translation assay. The assay showed that the proteins could be correctly translated from the plasmids and that the coNS3/4A plasmid gave detectable NS3 and NS3/4A bands at a higher plasmid dilution as compared to the wtNS3/4A plasmid (Figure 1). This suggests that the in vitro translation from the coNS3/4A plasmid might be more effective. To compare the expression levels more precisely, HepG2 cells were transiently transfected with the wtNS3/4A and the coNS3/4A plasmids. These experiments revealed that the coNS3/4A plasmid generated 11-fold higher expression levels of the NS3 protein when compared to the wtNS3/4A plasmid, as determined by densitometry and a standard curve of recombinant NS3 (rNS3; Figure 1 and data not shown). Since the wtNS3/4A and the coNS3/4A plasmids are identical in size, it is unlikely that there would be any major differences in transfection efficiencies between the plasmids. The enhanced expression of the wtNS3 gene seen when including NS4A in the SFV replicon system has been reported previously.23 Staining of coNS3/4A plasmid-transfected, and SFVinfected, BHK cells revealed a similar perinuclear and cytoplasmic distribution of the NS3 as previously observed, confirming an unchanged subcellular localization (Figure 1).13,23 Effect of codon optimization and mRNA amplification on humoral responses To test the intrinsic immunogenicity of the different NS3 genes, groups of BALB/c (H-2d) mice were immunized with the wtNS3/4A, coNS3/4A DNAs, or wtNS3/4A-

Figure 1 Analysis of the translation products from the plasmids wtNS3/ 4A-pVAX1 and coNS3/4A-pVAX1 by in vitro translation in the presence of 35S-methionine and SDS-PAGE electrophoresis (a). (a) Lanes 1 and 10 show the molecular weight marker (CFA 756; Amersham Pharmacia Biotech), lane 2 the 61 kDa kit control, lane 3 the negative control, lanes 4 and 5, 6.4 ng of coNS3/4A-pVAX1 and wtNS3/4A-pVAX1, respectively, lanes 6 and 7, 2.6 ng of coNS3/4A-pVAX1 and wtNS3/4A-pVAX1, respectively, and lanes 8 and 9, 1.0 ng of coNS3/4A-pVAX1 and wtNS3/ 4A-pVAX1, respectively. (b) Western blot analysis of immunoprecipitated HepG2 cells transiently transfected with wtNS3/4A (lane 1) and coNS3/ 4A (lane 2). The controls were recombinant NS3 (lane 3) and negative control plasmid (lane 4). (c) Immunofluorescent staining of BHK cells using an NS3-specific monoclonal antibody, 24 h after infection with wtNS3/4A-SFV or transfection with the coNS3/4A-pVAX1 plasmid.

SFV vectors. Doses of 4 mg DNA was administered using the gene gun and doses of 107 SFV particles were injected subcutaneously (s.c.). The mice were boosted after 4 weeks. The mice immunized with the wtNS3/4A-SFV developed antibodies already after the first injection suggesting a potent immunogenicity (Figure 2). At 2 weeks after the second immunization, most mice immunized with the coNS3/4A or wtNS3/4A-SFV vectors had developed mean antibody levels over 103 (Figure 2). In contrast, none of the mice immunized with the wtNS3/4A plasmid had developed detectable NS3specific antibodies at 6 weeks (Figure 2). Thus, both codon optimization and mRNA amplification by SFV result in an increased B-cell immunogenicity of the NS3/ 4A gene. The Th-cell phenotype primed by wtNS3/4A immunization has been described in detail previously.13,23 Gene Therapy


524

tion. When genetically immunizing BALB/c mice with NS3/4A using the gene gun, the subclass ratio suggested a mixed Th1/Th2 response (Figure 2). This is fully consistent with previous reports suggesting a less Th1like response primed by gene gun immunization.28 It should be noted that the codon-optimized plasmid did not display an increased in vitro stimulatory capacity of B cells, as compared to the native plasmid (data not shown), suggesting that no major immune stimulatory motifs had been lost or introduced. Immunizations using SFV primed a Th1- or mixed Th1/Th2-like isotype distribution. The anti-NS3 IgG2a/ IgG1 ratio following wtNS3/4A-SFV immunization ranged from 2.4 to 20 between different experiments (Figure 2 and data not shown), suggesting a Th1-like response. This is similar to the previous experience with SFV vectors where a Th1-skewed IgG subclass distribution has been observed.29

Figure 2 Mean NS3-specific antibody responses primed by gene gun immunizations with 4 mg wtNS3/4A-pVAX1 and coNS3/4A-pVAX1, or s.c. injection of 107 wtNS3/4A-SFV particles in groups of 10 H-2d mice (a). All mice were immunized at weeks 0 and 4. Values are given as mean end point antibody titers (7s.d.). (b) The IgG subclass patterns from groups of five mice immunized twice with wtNS3/4A-pVAX1 given i.m., coNS3/4A-pVAX1 given i.m. or by gene gun (gg), and wtNS3/4A-SFV given s.c. Values are given as mean end point antibody titers (7s.d.). The ‘**’ sign indicates a statistical difference of Po0.01, the ‘*’ sign indicates a difference of Po0.05, and NS (not significant) indicates no statistical difference (Mann–Whitney). The titer ratio obtained by dividing the mean end point titer of IgG2a antibodies to NS3 by the mean end point titer of IgG1 antibodies to NS3 is also given. A high ratio (43) indicates a Th1like response and a low ratio (o0.3) indicates a Th2-like response, whereas values within a three-fold difference from 1 (0.3–3) indicate a mixed Th1/ Th2 response.

To compare indirectly the T helper 1 (Th1) and Th2 skewing of the T-cell response primed by wtNS3/4A, coNS3/4A, and wtNS3/4A-SFV immunizations, the levels of NS3-specific IgG1 (Th2) and IgG2a (Th1) antibodies were analyzed (Figure 2). The IgG2a/IgG1 ratio in mice immunized with rNS3 with or without adjuvant was always o1 regardless of the murine haplotype,27 signaling a Th2-dominated response.28 In contrast, mice immunized intramuscularly (i.m.) with the wtNS3-, wtNS3/4A-, or coNS3/4A-containing plasmids had Th1-skewed Th-cell responses evidenced by IgG2a/IgG1 ratios of 41 (Figure 2).23 Thus, codon optimization did not change the IgG subclass distribuGene Therapy

Effect of codon optimization and mRNA amplification on priming NS3-specific cytotoxic T lymphocytes We first estimated the frequency of NS3-specific cytotoxic T lymphocytes (CTLs) that could be primed by gene gun immunization using the wtNS3-, wtNS3/4A-, and coNS3/4A-expressing plasmids. The coNS3/4A plasmid primed higher precursor frequencies of NS3-specific CTL as compared to the wtNS3 gene enforcing the importance of NS4A (Figure 3). No statistical difference in CTL precursor frequencies was noted between the wtNS3/ 4A- and coNS3/4A-expressing plasmids when analyzed directly ex vivo (Figure 3). A single immunization with the coNS3/4A plasmid or wtNS3/4A-SFV primed around 1% of peptide-specific CTLs within 2 weeks from immunization (Figure 3). The specificity of the detection of NS3-specific CTLs was confirmed by a 5-day restimulation in vitro with the NS3 peptide, by which high precursor frequencies were observed after immunization with the coNS3/4A gene (Figure 3). To compare directly the in vitro lytic activity of the NS3-specific CTLs primed by different vectors, a standard 51Cr-release assay was performed after one or two immunizations. The lytic activity of the in vivo primed CTLs was assayed on both NS3-peptide-loaded H-2Db-expressing RMA-S cells and EL-4 cells stably expressing NS3/4A. After one dose, the coNS3/4A plasmid and the wtNS3/4A-SFV vector were clearly more efficient than the wtNS3/4A plasmid in priming CTLs that lysed NS3-peptide-coated target cells (Figure 4). Thus, the CTL priming event was enhanced by codon optimization or mRNA amplification of the NS3/4A gene. The difference was less clear when using the NS3/ 4A-expressing EL-4 cells presumably since this assay is less sensitive (Figure 4). After two immunizations, all NS3/4A vectors seemed to prime NS3-specific CTLs with a similar efficiency (Figure 4). However, two immunizations with any of the NS3/4A-containing vectors were clearly more efficient in priming NS3-specific CTLs as compared to the plasmid containing only the wtNS3 gene (Figure 4), which is fully consistent with the CTL precursor analysis and previous observations.23 In conclusion, codon optimization or mRNA amplification of the NS3/4A gene seems to allow for a more rapid priming of NS3-specific CTLs. It should be noted that the responses were measured to what seems to be the


525

Figure 3 Flow cytometric quantification of the precursor frequency of NS3/4A-specific CD8 þ T cells using peptide-loaded H-2Db:Ig fusion protein. (a) The mean % NS3-specific CD8 þ T cells from groups of five mice immunized twice with wtNS3-pVAX1, wtNS3/4A-pVAX1, or coNS3/4A-pVAX1 using gene gun. The ‘*’ sign indicates a difference of Po0.05, and NS (not significant) indicates no statistical difference (Mann–Whitney). Also shown are the raw data from representative individual mice from the groups listed above (e, f, and h), as well as from individual mice immunized once with coNS3/4ApVAX1 (b) or wtNS3/4A-SFV (c). (d, g) Nonimmunized control mice from the different experiments. (i, j) The splenocytes were restimulated for 5 days with the NS3 peptides prior to analysis. A total of 150 000–200 000 data points were collected and the percentages of CD8 þ cells stained for H-2Db:Ig are indicated in parentheses in each dot-plot.

immunodominant H-2b-restricted NS3-derived CTL peptide,23 and it cannot be excluded that responses to other subdominant epitopes on NS3 may behave differently. However, several attempts to identify additional H-2brestricted NS3-derived CTL epitopes have thus far failed (unpublished data).

Characterization of the tumor inhibition model The hepatitis C virus is as far as we know today a noncytolytic virus. Therefore, a relevant model to characterize in vivo functional HCV-specific immune

responses is the determination of inhibition of tumor growth in vivo in BALB/c mice using SP2/0 myeloma cells, or in C57BL/6 mice using EL-4 lymphoma cells, expressing the desired viral antigen.8 An SP2/0 cell line stably expressing NS3/4A has been described previously 23 and we now established an NS3/4A-expressing EL-4 cell line. To confirm that inhibition of tumor growth using the NS3/4A-expressing EL-4 cell line is fully dependent on an NS3/4A-specific immune response, a control experiment was performed. Groups of 10 C57BL/6 mice were Gene Therapy


526

Figure 4 Priming of in vitro detectable CTLs in H-2b mice by gene gun immunization of the wtNS3-pVAX1, wtNS3/4A, and coNS3/4A plasmids, or s.c. injection of wtNS3/4A-SFV particles. Groups of 5–10 H-2b mice were immunized once (a) or twice (b). The percent specific lysis corresponds to the percent lysis obtained with either NS3-peptide-coated RMA-S cells (upper panel in (a, b)) or NS3/4A-expressing EL-4 cells (lower panel in (a, b)) minus the percent lysis obtained with unloaded or nontransfected EL-4 cells. Values are given for effector to target (E:T) cell ratios of 60:1, 20:1, and 7:1. Each line indicates an individual mouse.

either left nonimmunized, or immunized twice with the coNS3/4A plasmid. At 2 weeks after the last immunization, the mice were challenged with an s.c. injection of 106 native EL-4 or NS3/4A-expressing EL-4 cells (NS3/ 4A-EL-4). An NS3/4A-specific immune response was required for protection, since only the immunized mice were protected against growth of the NS3/4A-EL-4 cell line (Figure 5). Thus, this H-2b-restricted model behaves like the previously described H-2d-restricted model.23 Our previous data using immunizations with recombinant NS3 protein suggested that neither NS3/4Aspecific B cells nor CD4 þ T cells were of a pivotal importance in protection against tumor growth.23 In vitro depletion of CD4 þ or CD8 þ T cells of splenocytes from coNS3/4A plasmid-immunized H-2b mice suggested that CD8 þ T cells were the major effector cells in the 51 Cr-release assay (data not shown). To define the in vivo functional antitumor effector cell population, we selectively depleted CD4 þ or CD8 þ T cells in mice immunized with the coNS3/4A plasmid 1 week prior to, and during, challenge with the NS3/4A-EL-4 tumor cell line. Analysis by flow cytometry revealed that 85% of CD4 þ and CD8 þ T cells had been depleted (data not shown). This experiment revealed that in vivo depletion of CD4 þ T cells had no significant effect on the tumor immunity (Figure 5). In contrast, depletion of CD8 þ T cells in vivo significantly reduced the tumor immunity Gene Therapy

Figure 5 Specificity of tumor-inhibiting immune responses primed by gene gun (a). Groups of 10 C57BL/6 mice were either left untreated or were given two monthly immunizations with 4 mg of coNS3/4A-pVAX1. At 2 weeks after the last immunization, mice were injected subcutaneously with the parental EL-4 cell line or 106 NS3/4A-expressing EL-4 cells. Tumor sizes were measured through the skin at days 6, 7, 10, 11, 12, and 14 after tumor injection. (b) The in vivo functional effector cell population was determined in groups of 10 C57BL/6 mice immunized twice with the coNS3/4A-pVAX1 plasmid using gene gun. In two groups, either CD4 þ or CD8 þ T cells were depleted by administration of monoclonal antibodies 1 week prior to, and during, challenge with the NS3/4A-expressing EL-4 cell line. Tumor sizes were measured through the skin at days 5, 6, 8, 11, 13, 14, and 15 after tumor injection. Values are given as the mean tumor size7s.e. The ‘**’ sign indicates a statistical difference of Po0.01, the ‘*’ sign indicates a difference of Po0.05, and NS (not significant) indicates no statistical difference (area under the curve values compared by ANOVA).

(Po0.05, ANOVA; Figure 5). Thus, as expected, NS3/4Aspecific CD8 þ CTLs seem to be the major protective cells at the effector stage in the in vivo model for inhibition of tumor growth.

Effect of codon optimization and mRNA amplification on priming in vivo functional NS3-specific CTLs The tumor challenge model was used to evaluate how effective the different immunogens were in priming a protective immunity against growth of NS3/4A-EL-4 tumor cells in vivo. To ensure that the effectiveness of the priming event was studied, all mice were immunized only once. Fully consistent with the in vitro CTL data, we found that only vectors containing NS3/4A were able to prime rapidly protective immune responses as compared to immunized with the empty pVAX plasmid (Po0.05, ANOVA; Figure 6). However, this was dependent on NS4A but independent of either codon optimization or mRNA amplification, suggesting that C57BL/6 mice are quite easily protected against tumor growth using genetic immunization. To further clarify the prerequisites for priming of the in vivo protective CD8 þ CTL responses, additional experiments were performed. First, C57BL/6 mice immunized with the NS3-derived CTL peptide were not protected against growth of NS3/4A-EL-4 tumors


527

Figure 6 Evaluation of the ability of different immunogens to prime HCV NS3/4A-specific tumor-inhibiting responses after a single immunization. Groups of 10 C57BL/6 mice were either left untreated or were given one immunization with the indicated immunogen (4 mg DNA using gene gun in (a, b, c, g, and h); 107 SFV particles s.c. in (d); 100 mg peptide in CFA s.c. in (e); and 20 mg rNS3 in CFA s.c. in (f)). At 2 weeks after the last immunization, mice were injected subcutaneously with 106 NS3/4A-expressing EL-4 cells. Tumor sizes were measured through the skin at days 6–19 after tumor injection. Values are given as the mean tumor size7s.e. In (a–e), as a negative control the mean data from the group immunized with the empty pVAX plasmid by gene gun have been plotted in each graph. In (f–h), the negative controls were nonimmunized mice. Also given is the P-value obtained from the statistical comparison of the control with each curve using the area under the curve and ANOVA.

(Figure 6). Second, immunization with recombinant NS3 in adjuvant did not protect against tumor growth (Figure 6). We do know from previous studies that the NS3derived CTL peptide effectively primes CTLs in C57BL/ 6 mice and that rNS3 in adjuvant primes high levels of NS3-specific T helper cells. Thus, an endogenous production of NS3/4A seems to be needed to prime in vivo protective CTLs. To further characterize the priming event, groups of B-cell- (mMT30) or CD4-deficient C57BL/6 mice31 were immunized once with the coNS3/ 4A gene using gene gun, and were challenged 2 weeks later (Figure 6). Since both lineages were protected against tumor growth, we conclude that neither B cells nor CD4 þ T cells were required for the priming of in vivo functional NS3/4A-specific CTLs (Figure 6). In conclusion, the priming of in vivo tumor protective NS3/4A-specific CTLs in C57BL/6 mice requires NS4A and an endogenous expression of the immunogen. In C57BL/6 mice, the priming is less dependent on the gene delivery route or accessory cells, such as B cells or CD4 þ T cells. The fact that the priming of in vivo functional CTL by the coNS3/4A DNA plasmid was independent of CD4 þ T helper cells may help to explain the speed by which the priming occurred. Repeated experiments in C57BL/6 mice using the NS3/4A-EL-4 cell line have shown that protection against tumor growth is obtained already after the first immunization with the NS3/4A gene, independent of codon optimization or mRNA amplification (Figure 6 and data not shown). Also, after two injections, the immunity against NS3/4A-EL-4 tumor growth was even further enhanced, but only when NS4A was present (data not shown). Thus, this model may therefore not be sufficiently sensitive to reveal subtle differences in the intrinsic immunogenicity of different immunogens. Thus, to better compare the immunogenicity of the

wtNS3/4A and the coNS3/4A DNA plasmids, additional experiments were performed in H-2d mice, where at least two immunizations seemed to be required for a tumor protective immunity23 (data not shown). It is important to remember that the IgG subclass distribution obtained after gene gun immunization with the NS3/4A gene in BALB/c mice suggested a mixed Th1/Th2-like response. Thus, it was possible that a Th2-like immunization route (gene gun) in the Th2-prone BALB/c mouse strain32,33 may impair the ability to prime in vivo effective CTL responses. Groups of 10 BALB/c mice were immunized once, twice, or thrice with 4 mg of the respective DNA plasmid using the gene gun (Figure 7). The mice were challenged 2 weeks after the last injection. From these experiments, it became clear that the coNS3/ 4A plasmid primed an in vivo functional NS3/4Aspecific tumor-inhibiting immunity more rapidly than the wild-type plasmid (Figure 7). Two doses of the coNS3/4A primed a significantly better NS3/4A-specific tumor-inhibiting immunity as compared to the wtNS3/ 4A plasmid (Po0.05, ANOVA; Figure 7). After three doses, the tumor-inhibiting immunity was the same. Thus, we could confirm from the data from the in vitro experiments that codon optimization of the NS3/4A gene primes NS3-specific CTLs more rapidly. The primary aim with the NS3/4A vaccine is its use as a therapeutic vaccine. Since we now had at least one vaccine candidate that was well characterized and that quickly primed in vivo functional CTLs, we evaluated the effect of therapeutic immunization after the injection of tumor cells. Groups of 10 C57BL/6 mice were challenged with 106 NS3/4A-EL-4 tumor cells. One group was immunized transdermally with of 4 mg coNS3/4A 6 six days, and another group at 12 days, after tumor challenge. After the therapeutic vaccination, both groups had significantly smaller tumors as compared to the Gene Therapy


528

Figure 7 Comparative efficiency between the wtNS3/4A-pVAX1 and coNS3/4A-pVAX1 plasmids delivered by gene gun in priming tumor-inhibiting immune responses. Groups of 10 BALB/c mice were either left untreated or were given one, two, or three monthly immunizations with 4 mg of plasmid. At 2 weeks after the last immunization, mice were injected subcutaneously with 106 NS3/4A-expressing SP2/0 cells. Tumor sizes were measured through the skin at days 6, 8, 10, 11, 12, 13, and 14 after tumor injection. Values are given as the mean tumor size7s.e. The ‘**’ sign indicates a statistical difference of Po0.01, the ‘*’ sign indicates a difference of Po0.05, and NS (not significant) indicates no statistical difference (area under the curve values compared by ANOVA).

the coNS3/4A plasmid also works as a therapeutic vaccine.

Discussion

Figure 8 Effect of therapeutic vaccination with the coNS3/4A plasmid using the gene gun. Groups of 10 C57BL/6 mice were inoculated with 106 NS3/4A-EL4 cells. One group had been immunized once with 4 mg coNS3/ 4A DNA using gene gun 2 weeks prior to challenge (positive control), one group was immunized the same way 6 days after tumor inoculation, and one group was immunized 12 days after tumor inoculation. One group was not immunized (negative control). Tumor sizes were measured through the skin at days 6, 10, 11, 12, 13, 14, 18, 19, and 20 after tumor injection. Values are given as the mean tumor size7s.e. The ‘**’ sign indicates a statistical difference of Po0.01, the ‘*’ sign indicates a difference of Po0.05, and NS (not significant) indicates no statistical difference (area under the curve values compared by ANOVA).

nonimmunized control group (Po0.01, ANOVA; Figure 8). This confirms that the vaccine rapidly primes CTLs, which are able to home to and infiltrate the NS3/4Aexpressing tumors. Thus, gene gun immunization with Gene Therapy

Key factors in the design of genetic vaccines is to understand and enhance the intrinsic immunogenicity of the selected gene and to test which approach is the most suitable for the particular gene of interest. With respect to the NS3 protein of HCV, which has been evaluated in many laboratories as a therapeutic vaccine candidate,7,8,13,23,34,35 some important findings have recently been made. The NS3 protein has three enzymatic functions in the viral life cycle. This is the most probable reason for the genetic stability of the gene, which is of vital importance when considering the plasticity of the HCV genome in vaccine development. It was recently shown that the NS3/4A protein interferes in the cellular interferon response pathways,24 which suggests that the NS3/4A complex may play a key role in maintaining the chronic infection. We recently showed that the inclusion of NS4A in NS3-based genetic vaccines greatly enhances the expression levels and immunogenicity of NS3,23 which now has been further confirmed in the present report. Thus, the importance of NS4A in the pathogenicity of the virus may previously have been underestimated. It may be argued that the ability of NS3/4A to interfere with the cellular interferon signaling pathways could decrease the immunogenicity of the NS3/4A complex. However, this may not be true when expressing the gene outside the context of the natural infection. For example, DNA plasmids expressing NS3/4A are superior to plasmids expressing only NS3 with respect to the ability to prime both humoral and cellular responses.23 Certain CpG motifs may induce an increased production of IFN-alpha,36–38 a step that possibly could be inhibited by the presence of NS3/4A. Regardless of the effect, the complete NS3/4A complex is more immunogenic than


NS3 alone, suggesting that the presence of NS4A has more immune enhancing than inhibiting effects. In SFVinfected cells, the replicase produces double-stranded (ds) RNA when synthesizing the viral mRNA.39 Also, the SFV replication causes the infected cell to undergo apoptosis.39 Most likely due to the cellular recognition of dsRNA as a pathogen-associated molecular pattern (PAMP),40 a potent interferon (IFN) response is activated in SFV-infected cells.41 The activation of the cellular IFN system is intended to inhibit the SFV-mediated protein expression. However, if the NS3/4A complex is present, it is possible that the interferon signaling pathways activated by the presence of dsRNA are inhibited. Further studies are needed using the SFV system with NS3 alone and the combination of NS3 and NS4A to understand fully the effects that modulating the IFN system has on immunogenicity. As we show herein, the combination of NS3/4A and delivery by SFV clearly improves the immunogenicity of the NS3/4A gene. It is therefore a real possibility that both the inhibition of the cellular interferon response and the mRNA amplification contribute to the potent immunogenicity of the NS3/4ASFV combination. There are several ways to improve the immunogenicity of genetic vaccines, such as co-administration of various cytokines or cytokine genes and the inclusion of immune stimulatory CpG motifs.37 However, the first step should include optimizing the intrinsic immunogenicity of the gene as much as possible. With respect to the NS3 gene of HCV, the first step was to include NS4A, which was clearly beneficial. We next sought of ways to further improve the immunogenicity by modulating the protein expression levels. We could show that both codon optimization and mRNA amplification are highly effective in enhancing the immunogenicity of the NS3/ 4A complex. Importantly, codon optimization clearly improved both the priming of humoral responses and specific CTLs. The CTL responses appeared more rapidly when immunizing with the codon-optimized NS3/4A gene as compared to the wild-type gene. In addition, gene gun immunization with the coNS3/4A gene given as a therapeutic vaccine 6–12 days after inoculation with tumor cells slowed the rate of tumor growth. Thus, this confirms the rapid priming of in vivo functional NS3/4Aspecific CTLs using the codon-optimized gene. The approach of mRNA amplification of the wild-type NS3/4A gene using the SFV replicon revealed a superior B-cell immunogenicity and a more rapid induction of in vitro-detectable CTLs, as compared to the wild-type NS3/4A gene delivered as a DNA immunogen. These data strongly suggest that expression levels, and/or the mode of expression, are of key importance also in an effective priming of CTLs. We could show that the priming of a protective immunity against growth of NS3/4A-expressing tumor cells requires both an endogenous expression of the immunogen and CD8 þ T cells. Thus, the protection seems to be mainly mediated by CD8 þ CTLs primed through the endogenous antigen-presentation pathway. We note that the priming of NS3/4A-specific CD8 þ CTLs using gene gun was independent of the presence of B and CD4 þ T helper cells. This is consistent with a previous report showing that the priming of CTLs using gene gun occurred through crosspriming and was independent of CD4 þ T-cell help.42 However, the need

for CD4 þ help in the priming of CTLs is still controversial, whereas recent data suggest that CD4 þ T help is needed for maintenance of a long-lived CTL response.43 We have shown that the immunogenicity of the NS3 gene can be further improved by inclusion of NS4A. What we know today is that NS4A, by a not yet known mechanism, enhances the expression of NS3 in vitro and its immunogenicity in vivo. Recent data suggest that the NS3/4A complex can inhibit IFN-alpha/beta signaling,24,44 but its role in the immunogenicity of NS3/4A is not known. We showed that the immunogenicity of the NS3/4A gene could be even further improved by modulating translation or mRNA expression. We also showed that the priming of in vivo functional CTLs that prevent growth of NS3/4A-expressing tumor cells requires endogenous expression of the immunogen but is independent of CD4 þ T help. The latter most likely explains why the priming of CTLs is rapid. The rapid priming of CTLs using the codon-optimized gene proved to be of value in a therapeutic vaccination regimen. Gene gun immunization using the coNS3/4A gene 6–12 days after inoculation of NS3/4A-expressing tumor cells significantly inhibited tumor growth. Overall, a rapid priming of HCV NS3-specific immune responses that are functional in vivo is generated by either DNA-based immunization with a codon-optimized gene or by mRNA amplification by the SFV replicon. These approaches are applicable for development of a therapeutic vaccine for chronic HCV infections.

529

Materials and methods Mice Inbred BALB/c (H-2d) and C57BL/6 (H-2b) mice were obtained from commercial vendors (Mo¨llegard, Denmark). B-cell- (mMT30) deficient mice were kindly provided by Dr Karin Sandstedt, Karolinska Institutet, Sweden. CD4-deficient C57BL/6 mice31 mice were obtained from the breeding facility at the Microbiology and Tumourbiology Centre, Karolinska Institutet. The ethical committee for animal research at Karolinska Institutet had approved all animal experiments. All mice were female and were used at 4–8 weeks of age at the start of the experiments. Synthetic peptides The nine-mer peptide (sequence GAVQNEVTL), corresponding to an NS3 CTL epitope in H-2b mice, was synthesized by automated peptide synthesis as described previously.23 Recombinant NS3 ATPase/helicase domain protein The recombinant NS3 (rNS3) protein was kindly provided by Darrell L Peterson, Department of Biochemistry, Commonwealth University, VA, USA. The production of recombinant NS3 protein (not including NS4A) in Escherichia coli has been described in detail previously.45 Prior to use, the rNS3 protein was dialyzed overnight against phosphate-buffered saline (PBS) and sterile-filtered. Gene Therapy


530

Generation of a synthetic codon-optimized (co) NS3/4A gene The sequence of the previously isolated and sequenced unique wtNS3/4A gene, recently described,23 was analyzed for codon usage with respect to the most commonly used codons in human cells. A total of 435 nucleotides were replaced to optimize codon usage for human cells. The sequence was sent to Retrogen Inc. (San Diego, CA, USA) for generation of a full-length synthetic coNS3/4A gene. The coNS3/4A gene had a sequence homology of 79% with the region at nucleotide positions 3417–5475 of the HCV-1 reference strain.46 A total of 433 nucleotides differed. On an amino-acid level, the homology with the HCV-1 strain was 98% (15 amino acids differed). The full-length codon optimized 2.1 kb DNA fragment of the HCV genotype 1b corresponding to amino acids 1007–1711 encompassing NS3 and NS4A. NS3/NS4A gene fragment was inserted into a BamHI- and XbaIdigested pVAX vector (Invitrogen, San Diego, CA, USA) to give the coNS3/4A-pVAX plasmid. The expression construct was sequenced to ensure correct sequence and reading frame. The protein expression was analyzed by an in vitro transcription and translation assay. Plasmids were grown in competent TOP10 E. coli (Invitrogen). Plasmid DNA used for in vivo injection was purified by using Qiagen DNA purification columns according to the manufacturer’s instructions (Qiagen GmbH, Hilden, FRG). The concentration of the resulting plasmid DNA was determined spectrophotometrically (Dynaquant, Pharmacia Biotech, Uppsala, Sweden). Purified DNA was dissolved in sterile PBS at concentrations of 1 mg/ml. In vitro translation assay To ensure that the wtNS3/4A and coNS3/4A genes were intact and could be translated, an in vitro transcription assay using the procaryotic T7 coupled reticulocyte lysate system (TNT; Promega, Madison, WI, USA) was performed as described previously.13,23 To compare the translation efficiency of the two plasmids, the amount of input DNA was diluted in serial dilutions (6 to 1 ng) prior to the TNT assay. Transient transfections HepG2 cells were transiently transfected by standard protocols. In brief, HepG2 cells were plated into 2.5 cm2 wells (0.5 106) in DMEM medium the day before transfection. In all, 2 mg of each plasmid DNA construct (wtNS3/4A and coNS3/4A) was transfected into HepG2 cells using Fugene 6 Transfection Reagent (Roche). After transfection, the HepG2 cells were incubated for 24–48 h. Protein sample preparation and analysis Cell lysates were analyzed by immunoprecipitation followed by SDS-PAGE. In brief, transient transfected HepG2 cells were lysed in RIPA buffer (0.15 M NaCl, 50 mM Tris, 1% Triton X-100, 1% Na-deoxycholate, and 1% SDS). The cell lysates were immunoprecipitated with protein A sepharose and anti-NS3 polyclonal antibody overnight at 41C. The washed pellets were resuspended in SDS sample buffer, heated at 1001C for 5 min prior to SDS-PAGE analysis on 4–12% Bis-Tris gel (Invitrogen) and electrotransferred onto Nitrocellulose membranes. Gene Therapy

Analysis of NS3 protein expression Detection of NS3 protein was performed according to the manufacturer’s protocol by using a chemiluminescencelinked Western blot kit (WesternBreeze; Invitrogen). NS3 protein expression was detected and quantified as a chemiluminescent signal by using an NS3-specific polyclonal antibody. Chemiluminescent signals were detected by using the GeneGnome (Syngene, Cambridge, UK). Quantification of chemiluminescence Western blots was performed on GeneGnome, and units of intensity from each protein band were calculated and compared to a standard curve of rNS3. Semliki forest virus vectors Baby hamster kidney (BHK)-21 cells were maintained in complete BHK medium supplemented with 5% fetal calf serum (FCS), 10% tryptose phosphate broth, 2 mM glutamine, 20 mM Hepes, and antibiotics (streptomycin 10 mg/ml and penicillin 100 IU/ml). The sequence encoding wtNS3/4A has been described previously.23 The wtNS3/4A gene was isolated by PCR as Spe1-BStB1 fragment and inserted into the Spe1-BstB1 site of pSFV10Enh containing a 34 amino-acid long translational enhancer sequence of capsid followed by the FMDV 2a cleavage peptide.47,48 Packaging of recombinant RNA into rSFV particles was carried out using a two-helper RNA system.47 Indirect immunofluorescence of infected BHK cells was performed to determine the titer of the recombinant virus stocks.49 Immunofluorescence BHK cells were transiently transfected with coNS3/4ApVAX1 according to standard techniques using Lipofectamine plus reagent (Invitrogen) or infected by rSFV as described previously.23 NS3 protein was detected by indirect immunofluorescence.23 Immunization protocols Groups (5–10 mice/group) of female BALB/c (H-2d) or C57BL/6 (H-2b) mice, 4–8 weeks old, were immunized by needle injections of 100 mg of plasmid DNA encoding individual or multiple HCV proteins. Plasmid DNA in PBS was given i.m. in the tibialis anterior (TA) muscle.50 Where indicated in the text, the mice were injected i.m. with 50 ml/TA of 0.01 mM cardiotoxin (Latoxan, Rosans, France) in 0.9% sterile saline NaCl, 5 days prior to DNA immunization. The mice were boosted at 4-week intervals. For gene gun-based immunizations, plasmid DNA was linked to gold particles (1 mm) according to protocols supplied by the manufacturer (Bio-Rad Laboratories, Hercules, CA, USA). Prior to immunization, the abdominal injection area was shaved and the immunization was performed according to the manufacturer’s protocol at a helium discharge pressure of 500 psi. Each injection dose contained 4 mg of plasmid DNA. The mice were boosted with the same dose at monthly intervals. For rSFV particle immunizations, mice were immunized subcutaneously, in the base of the tail, with 1 107 virus particles diluted in PBS (wtNS3/4A-SFV), in a final volume of 100 ml. Peptide immunization was performed by subcutaneous immunization in the base of the tail with 100 mg peptide mixed 1:1 in complete Freunds adjuvant.


ELISA for detection of murine anti-HCV NS3 antibodies Serum for antibody detection and isotyping was collected every second or fourth week after the first immunization by retro-orbital bleeding of isofluoraneanesthetized mice. The enzyme immunoassays were performed as described previously.13,27 Cell lines The SP2/0-Ag14 myeloma cell line (H-2d) was maintained in DMEM medium supplemented with 10% FCS (Sigma Chemicals, St Louis, MO, USA), 2 mM L-glutamin, 10 mM HEPES, 100 U/ml penicillin, 100 mg/ml streptomycin, 1 mM nonessential amino acids, 50 mM b-mercaptoethanol, and 1 mM sodium pyruvate (GIBCO-BRL, Gaithersburgh, MD, USA). SP2/0-Ag14 cells with stable expression of NS3/4A were maintained in 800 mg geneticin (G418)/ml complete DMEM medium.23 The EL-4 lymphoma cell line (H-2b) was maintained in RPMI 1640 medium supplemented with 10% FCS, 10 mM HEPES, 1 mM sodium pyruvate, 1 mM nonessential amino acids, 50 mM b-mercaptoethanol, 100 U/ml penicillin, and 100 mg/ml streptomycin (GIBCO-BRL). EL-4 cells with stable expression of NS3/4A were generated by transfection of EL-4 cells with the linearized NS3/4ApcDNA3.1 plasmid using the SuperFect (Qiagen GmbH, Hilden, FRG) transfection reagent. The transfection procedure was performed according to the manufacturer’s protocol. Transfected cells were cloned by limiting dilution and were selected by addition of 800 mg geneticin (G418)/ml complete RPMI 1640 medium. RMA-S cells (a kind gift from Professor Klas Ka¨rre, Karolinska Institutet, Sweden) were maintained in RPMI 1640 medium supplemented with 5% FCS, 2 mM Lglutamin, 100 U/ml penicillin, and 100 mg/ml streptomycin. All cells were grown in a humidified 371C and 5% CO2 incubator. In vivo depletion of T cells CD4 and CD8 T-cell subpopulations were depleted in vivo by intraperitoneal injection of purified hybridoma supernatant. A total of 0.4 mg per mouse per injection of anti-CD4 (clone GK1.5) or anti-CD8 (clone 53–6.7) was injected on days 3, 2, and 1 before tumor challenge, and on days 3, 6, 10, and 13 after challenge. Flow cytometric analysis of peripheral blood mononuclear cell populations at days 0, 3, 6, 10, and 13 demonstrated that more than 85% of the CD4 and CD8 T cells were depleted. In vivo challenge with the NS3/4A-expressing tumor cells In vivo challenge of immunized mice with the NS3/4Aexpressing SP2/0 myeloma or EL-4 lymphoma cell line was performed according to the method described by Encke et al.8 In brief, groups of BALB/c or C57BL/6 mice were immunized with different immunogens at weeks 0, 4, and 8 as described. At 2 weeks after the last immunization, 1 106 NS3/4A-expressing SP2/0 or EL4 cells were injected subcutaneously in the right flank. The kinetics of the tumor growth was determined by measuring the tumor size through the skin at days 6–20. The kinetics of tumor development in two groups of mice was compared using the area under the curve

(AUC). The mean tumor sizes were compared using the analysis of variance (ANOVA) test. At day 20, all mice were killed. To test the therapeutic effect of the vaccines, groups of mice were inoculated with the tumor cells as described above. After 6 or 12 days, the mice were immunized once. The tumor growth was monitored from day 6 to day 20.

531

Antibodies and MHC:Ig fusion protein All monoclonal antibodies and MHC:Ig fusion proteins51 were purchased from BDB Pharmingen (San Diego, CA, USA): anti-CD16/CD32 (Fc-blockt, clone 2.4G2), FITCconjugated anti-CD8 (clone 53-6.7), Cy-Chrome-conjugated anti-CD4 (clone RM4-5), FITC-conjugated anti-H2Db (clone KH95), recombinant soluble dimeric mouse H-2Db:Ig, PE-conjugated rat-a mouse IgG1 (clone X56). Detection of NS3/4A-specific CTL activity Spleen cells from DNA- or rSFV-immunized C57BL/6 mice were resuspended in complete RPMI 1640 medium supplemented with 10% FCS, 2 mM L-glutamin, 10 mM HEPES, 100 U/ml penicillin and 100 mg/ml streptomycin, 1 mM nonessential amino acids, 50 mM b-mercaptoethanol, and 1 mM sodium pyruvate. In vitro stimulation was carried out for 5 days in 25-ml flasks at a final volume of 12 ml, containing 5 U/ml recombinant murine IL-2 (mIL-2; R&D Systems, Minneapolis, MN, USA). The restimulation culture contained a total of 25 106 immune spleen cells and 2.5 106 irradiated (10 000 rad) syngeneic EL-4 cells expressing the NS3/4A protein. After 5 days in vitro stimulation, a standard 51Crrelease assay was performed. Effector cells were harvested and a 4-h 51Cr assay was performed in 96-well U-bottom plates in a total volume of 200 ml. A total of 1 106 target cells (NS3/4A-expressing EL-4 cells) was labeled for 1 h at þ 371C with 20 ml of 51Cr (5 mCi/ml) and then washed three times in PBS. Different numbers of effectors and 51Cr-labeled target cells (5 103 cells/ well) were added to wells at effector:target (E:T) ratios of 60:1, 20:1, and 7:1. The level of cytolytic activity was determined after incubation of effectors and targets for 4 h at þ 371C. A volume of 100 ml supernatant was harvested and the radioactivity was measured with a g-counter. Splenocytes from DNA- or rSFV-immunized mice were harvested from C57BL/6 mice and were resuspended in complete RPMI 1640 medium as described previously.23 In brief, in vitro stimulation was carried out for 5 days by mixing 25 106 spleen cells and 25 106 irradiated (2000 rad) syngeneic splenocytes. The restimulation was performed in the presence of 0.05 mM NS3/ 4A H-2Db binding peptide (sequence GAVQNEVTL23). After restimulation, a 4-h 51Cr-release assay was performed using 51Cr-labeled peptide-pulsed RMA-S cells as targets. Cytotoxic activity was determined at the E:T ratios of 60:1, 20:1, and 7:1. Results were expressed according to the following formula: percent specific lysis ¼ (experimental releasespontaneous release)/(maximum releasespontaspontaneous release). Experimental release is the mean counts/minute released by the target cells in the presence of effector cells. Maximum release is the radioactivity released after lysis of target cells with 10% Gene Therapy


532

Triton X-100. Spontaneous release is the leakage of radioactivity into the medium of target cells. In vitro T-cell depletion experiments were conducted by incubating effector cells with either an anti-CD4 or anti-CD8, monoclonal antibody containing hybridoma supernatant (clone RL 172.4: anti-CD4; or clone 31M: anti-CD8) for 30 min at 41C. The cells were then washed and incubated at 371C for 1 h with complement (1/20 dilution of low-toxicity rabbit complement; Saxon, UK) before performing the CTL assay described above.

Quantification of NS3/4A-specific CTLs by flow cytometry The frequency of NS3-peptide-specific CD8 þ T cells was analyzed by ex vivo staining of spleen cells from DNA- or rSFV-immunized mice with recombinant soluble dimeric mouse H-2Db:Ig fusion protein as described previously.23,52 In brief, spleen cells were resuspended in PBS/1% FCS (FACS buffer) and incubated with Fcblocking antibodies. Cells were then washed and incubated with H-2Db:Ig preloaded with NS3/4Aderived peptide. Afterwards, cells were washed and incubated with PE-conjugated rat-a mouse IgG1 antibody, FITC-conjugated a-mouse CD8 antibody, and CyChrome a-mouse CD4 antibody. After washing, the cells were diluted in FACS buffer containing propidium iodide (PI). Approximately 200 000 total events from each sample were acquired on a FACSCalibur (BDB), and dead cells (PI-positive cells) were excluded in the analysis. Statistical analysis Fisher’s exact test was used for frequency analysis and Mann–Whitney U-test was used for comparing values from two groups. Kinetic tumor development in two groups of mice was compared using the AUC. AUC values were compared using analysis of variance (ANOVA). The calculations were performed using the Macintosh version of the StatView software (version 5.0).

Acknowledgements The study was supported by Grant no. K2000-06X-1261703A and K2002-16X-09494-12B from the Swedish Research Council, and by Grant no. QLK2-1999-00588 from the European Community.

References 1 Di Bisceglie AM, Hoofnagle JH. Optimal therapy of hepatitis C. Hepatology 2002; 36: S121–S127. 2 Weiner AJ et al. Evidence for immune selection of Hepatitis-C Virus (HCV) putative envelope glycoprotein variants – potential role in chronic HCV infections. Proc Natl Acad Sci USA 1992; 89: 3468–3472. 3 Ramratnam B et al. Rapid production and clearance of HIV-1 and hepatitis C virus assessed by large volume plasma apheresis. Lancet 1999; 354: 1782–1785. 4 Ogata N, Alter HJ, Miller RH, Purcell RH. Nucleotide sequence and mutation rate of the H strain of hepatitis C virus. Proc Natl Acad Sci USA 1991; 88: 3392–3396. 5 Lu L et al. Evaluation of accumulation of hepatitis C virus mutations in a chronically infected chimpanzee: comparison of Gene Therapy

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

the core, E1, HVR1, and NS5b regions. J Virol 2001; 75: 3004–3009. Bocher WO et al. Induction of strong hepatitis B virus (HBV) specific T helper cell and cytotoxic T lymphocyte responses by therapeutic vaccination in the trimera mouse model of chronic HBV infection. Eur J Immunol 2001; 31: 2071–2079. Brinster C et al. Different hepatitis C virus nonstructural protein 3 (Ns3)-DNA- expressing vaccines induce in HLA-A2.1 transgenic mice stable cytotoxic T lymphocytes that target one major epitope. Hepatology 2001; 34: 1206–1217. Encke J, zu Putlitz J, Geissler M, Wands JR. Genetic immunization generates cellular and humoral immune responses against the nonstructural proteins of the hepatitis C virus in a murine model. J Immunol 1998; 161: 4917–4923. Forns X et al. DNA immunization of mice and macaques with plasmids encoding hepatitis C virus envelope E2 protein expressed intracellularly and on the cell surface. Vaccine 1999; 17: 1992–2002. Geissler M et al. Differential cellular and humoral immune responses to HCV core and HBV envelope proteins after genetic immunizations using chimeric constructs. Vaccine 1998; 16: 857–867. Gordon EJ et al. Immune responses to hepatitis C virus structural and nonstructural proteins induced by plasmid DNA immunizations. J Infect Dis 2000; 181: 42–50. Inchauspe G et al. Plasmid DNA expressing a secreted or a nonsecreted form of hepatitis C virus nucleocapsid: comparative studies of antibody and T-helper responses following genetic immunization. DNA Cell Biol 1997; 16: 185–195. Lazdina U et al. Humoral and CD4(+) T helper (Th) cell responses to the hepatitis C virus non-structural 3 (NS3) protein: NS3 primes Th1-like responses more effectively as a DNA-based immunogen than as a recombinant protein. J Gen Virol 2001; 82: 1299–1308. Major ME et al. DNA-based immunization with chimeric vectors for the induction of immune responses against the hepatitis C virus nucleocapsid. J Virol 1995; 69: 5798–5805. Satoi J et al. Genetic immunization of wild-type and hepatitis C virus transgenic mice reveals a hierarchy of cellular immune response and tolerance induction against hepatitis C virus structural proteins. J Virol 2001; 75: 12121–12127. Tokushige K et al. Expression and immune response to hepatitis C virus core DNA-based vaccine constructs. Hepatology 1996; 24: 14–20. Bartenschlager R, Ahlborn-Laake L, Mous J, Jacobsen H. Nonstructural protein 3 of the hepatitis C virus encodes a serine-type proteinase required for cleavage at the NS3/4 and NS4/5 junctions. J Virol 1993; 67: 3835–3844. Bartenschlager R, Lohmann V, Wilkinson T, Koch JO. Complex formation between the NS3 serine-type proteinase of the hepatitis C virus and NS4A and its importance for polyprotein maturation. J Virol 1995; 69: 7519–7528. Failla C, Tomei L, De Francesco R. Both NS3 and NS4A are required for proteolytic processing of hepatitis C virus nonstructural proteins. J Virol 1994; 68: 3753–3760. Pang PS, Jankowsky E, Planet PJ, Pyle AM. The hepatitis C viral NS3 protein is a processive DNA helicase with cofactor enhanced RNA unwinding. EMBO J 2002; 21: 1168–1176. Wolk B et al. Subcellular localization, stability, and transcleavage competence of the hepatitis C virus NS3–NS4A complex expressed in tetracycline- regulated cell lines [In Process Citation]. J Virol 2000; 74: 2293–2304. Tanji Y et al. Hepatitis C virus-encoded nonstructural protein NS4A has versatile functions in viral protein processing. J Virol 1995; 69: 1575–1581. Frelin L et al. Low dose and gene gun immunization with a hepatitis C virus nonstructural (NS) 3 DNA-based vaccine


533 24 25

26

27

28

29

30

31

32

33

34

35

36

containing NS4A inhibit NS3/4A-expressing tumors in vivo. Gene Therapy 2003; 10: 686–699. Foy E et al. Regulation of interferon regulatory factor-3 by the Hepatitis C virus serine protease. Science 2003; 300: 1145–1148. Deml L et al. Multiple effects of codon usage optimization on expression and immunogenicity of DNA candidate vaccines encoding the human immunodeficiency virus type 1 Gag protein. J Virol 2001; 75: 10991–11001. Cid-Arregui A, Juarez V, zur Hausen H. A synthetic E7 gene of human papillomavirus type 16 that yields enhanced expression of the protein in mammalian cells and is useful for DNA immunization studies. J Virol 2003; 77: 4928–4937. Sallberg M et al. Immunogenicity and antigenicity of the ATPase/helicase domain of the hepatitis C virus nonstructural 3 protein. J Gen Virol 1996; 77: 2721–2728. Schirmbeck R, Reimann J. Modulation of gene-gun-mediated Th2 immunity to hepatitis B surface antigen by bacterial CpG motifs or IL-12. Intervirology 2001; 44: 115–123. Chen M et al. Vaccination with recombinant alphavirus or immune-stimulating complex antigen against respiratory syncytial virus. J Immunol 2002; 169: 3208–3216. Kitamura D, Roes J, Kuhn R, Rajewsky K. A B cell-deficient mouse by targeted disruption of the membrane exon of the immunoglobulin mu chain gene. Nature 1991; 350: 423–426. Rahemtulla A et al. Normal development and function of CD8+ cells but markedly decreased helper cell activity in mice lacking CD4. Nature 1991; 353: 180–184. Sadick MD et al. Cure of murine leishmaniasis with antiinterleukin 4 monoclonal antibody. Evidence for a T celldependent, interferon gamma-independent mechanism. J Exp Med 1990; 171: 115–127. Savelkoul HF, Termeulen J, Coffman RL, Van der Linde-Preesman RA. Frequency analysis of functional Ig C epsilon gene expression in the presence and absence of interleukin 4 in lipopolysaccharide-reactive murine B cells from high and low IgE responder strains. Eur J Immunol 1988; 18: 1209–1215. Jiao X et al. Modulation of cellular immune response against hepatitis C virus nonstructural protein 3 by cationic liposome encapsulated DNA immunization. Hepatology 2003; 37: 452–460. Brinster C et al. Hepatitis C virus non-structural protein 3specific cellular immune responses following single or combined immunization with DNA or recombinant Semliki Forest virus particles. J Gen Virol 2002; 83: 369–381. Cho HJ et al. IFN-alpha beta promote priming of antigen-specific CD8+ and CD4+ T lymphocytes by immunostimulatory DNAbased vaccines. J Immunol 2002; 168: 4907–4913.

37 Krieg AM. CpG motifs in bacterial DNA and their immune effects. Annu Rev Immunol 2002; 20: 709–760. 38 Krug A et al. Identification of CpG oligonucleotide sequences with high induction of IFN-alpha/beta in plasmacytoid dendritic cells. Eur J Immunol 2001; 31: 2154–2163. 39 Atkins GJ, Sheahan BJ, Liljestrom P. The molecular pathogenesis of Semliki Forest virus: a model virus made useful? J Gen Virol 1999; 80 (Part 9): 2287–2297. 40 Matzinger P. The danger model: a renewed sense of self. Science 2002; 296: 301–305. 41 Baigent SJ et al. Inhibition of beta interferon transcription by noncytopathogenic bovine viral diarrhea virus is through an interferon regulatory factor 3-dependent mechanism. J Virol 2002; 76: 8979–8988. 42 Cho JH, Youn JW, Sung YC. Cross-priming as a predominant mechanism for inducing CD8(+) T cell responses in gene gun DNA immunization. J Immunol 2001; 167: 5549–5557. 43 Janssen EM et al. CD4+ T cells are required for secondary expansion and memory in CD8+ T lymphocytes. Nature 2003; 421: 852–856. 44 Blindenbacher A et al. Expression of hepatitis c virus proteins inhibits interferon alpha signaling in the liver of transgenic mice. Gastroenterology 2003; 124: 1465–1475. 45 Jin L, Peterson DL. Expression, isolation, and characterization of the hepatitis C virus ATPase/RNA helicase. Arch Biochem Biophys 1995; 323: 47–53. 46 Choo QL et al. Genetic organization and diversity of the hepatitis C virus. Proc Natl Acad Sci USA 1991; 88: 2451–2455. 47 Smerdou C, Liljestrom P. Two-helper RNA system for production of recombinant Semliki forest virus particles. J Virol 1999; 73: 1092–1098. 48 Smerdou C, Liljestrom P. Non-viral amplification systems for gene transfer: vectors based on alphaviruses. Curr Opin Mol Ther 1999; 1: 244–251. 49 Liljestro¨m P, Garoff H. Expression of Proteins Using SemlikiForest Virus Vectors. In current Protocols in Molecular Biology. Greene Publishing Associates and Wiley Interscience, 1994, pp 1092–1098. 50 Davis HL et al. Plasmid DNA is superior to viral vectors for direct gene transfer into adult mouse skeletal muscle. Hum Gene Ther 1993; 4: 733–740. 51 Dal Porto J et al. A soluble divalent class I major histocompatibility complex molecule inhibits alloreactive T cells at nanomolar concentrations. Proc Natl Acad Sci USA 1993; 90: 6671–6675. 52 Lazdina U et al. Molecular basis for the interaction of the hepatitis B virus core antigen with the surface immunoglobulin receptor on naive B cells. J Virol 2001; 75: 6367–6374.

Gene Therapy